Scaling a Big Data / IoT Company Toward $1 Billion
October 11, 2016
Vital Lessons on Hyper-Growth from Splunk’s former CEO
At a recent session in San Francisco for a dozen CEOs of young tech companies in the investment portfolio of UK and New York based Octopus Ventures, Godfrey Sullivan, chairman and former CEO of Splunk, provided the group with some powerful insights about how he and his team managed to drive Splunk’s growth from $10m. in revenue in 2008, when he began in the company as CEO, to its current run-rate closing in on $1bn. eight years later. This period included a 2012 IPO that has turned Splunk into probably the single most successful big data public company with a valuation of $8bn. as of October 7. During the eight years the customer base grew from 750 to over 10,000 and today Splunk is an acknowledged leader in big data analytics, security, and internet of things (IoT) applications.
Sullivan retired in mid-November 2015, when sales were reaching the level of $650m annually, and the company’s revenues continue to grow at a 50% clip annually. The company is now led by Doug Merritt, a former SAP senior executive with prior leadership stints at Cisco and a couple of startups during the past 15 years. Sullivan brought Merritt into the company two and a half years ago to run worldwide field operations.
Here are some of the questions fielded by Sullivan. What the group heard from him exemplified a leadership philosophy founded on deep common sense, transparency, and fairness to all stakeholders – characteristics not always in plentiful supply in our industry. Sullivan is a respected and seasoned business leader with prior leadership roles at Hyperion and Autodesk. Many of his responses contained contrarian pearls of wisdom from which I believe every entrepreneur, CEO, executive, and board member can learn.
Below are the questions posed by the CEOs followed by Sullivan’s remarks. These are followed by my summary of the ten key lessons I took from this session.
What did you have to change when you arrived at Splunk?
“First let me address what the founders and their employees had done right: they had built a great product, providing a terrific customer experience, so there was a lot to like. On the negative side, we were losing money on every sale, burning through our C Round and running out of cash – this was 2008/9 and any funding would have meant heavy dilution. So we did a couple of things: we downsized the company in a brutal act of discipline that I referred to as our death march; and we doubled our prices for our entry level offering. This wasn’t popular at first with the founders nor others on the board. In fact, some of our investors would have preferred that we take additional funding, but at what cost to the company and its other shareholders?”
Did you market Splunk as a platform or as an application?
“I think of ‘platform’ as a suicidal term. When thinking about how to best position us, I thought it best to ask our sys admin customer ‘what do you call us?’ they generally said, ‘a tool – for IT systems management’. So I had the word ‘platform’ removed from all our literature for four years and we marketed our IT toolkit. When your customers say you’re a platform, then you can call yourself a platform. We only allowed the use of the word ‘platform’ in the past two years or so. Everyone wants to be a platform, but until your customer sees it as such, you’re solving problems and selling applications.”
How did you scale your sales operations?
“You have to have a repeatable sales and deployment motion. We started by building and marketing an IT operations tool to IT Ops buyers. Then we bought a company to add security as an app on top of our engine. I had to fight Gartner not to put us in the SIEM Magic Quadrant because we didn’t want to be trapped in a security positioning. That said, as we got into the security market, we had to hire security specialists so that we could gain credibility when addressing our customers’ security needs. Fortunately, selling to CISOs was a reasonably close adjacency to selling to IT Ops, even though it wasn’t the same buying center, but they were all in IT. Expanding to new market segments is a challenge for every company, and we try to find close adjacencies to our core.”
How else did you generate such consistently rapid growth?
“Splunk has the most passionate customers I’ve ever seen. Our go-to-market strategy was always centered on showcasing our customers. Today we hold around one hundred seminars each year around the world, and many of them have a thousand or more people subscribing, always with customer speakers. Technical users only believe other technical users.”
What was it that super-charged your growth?
“We have a freemium model that allows the tech user to satisfy their curiosity and need before they spend any money. This turned the traditional enterprise license pricing model upside down and we operate this model to this day.”
What do you think makes a good salesperson?
“Commitment to a process – do it the way the company wants you to do it, don’t fight it.”
What matters more in salespeople, IQ or EQ?
“EQ matters more, because customers need to like you, trust you, and appreciate what you do to help them be successful.”
How did you go about making your numbers every quarter? (Splunk has a track record of exceeding expectations quarter after quarter for the past four years since the IPO.)
“We are known for conservative guidance. We created what I call our Success Model. That is, we had no overlays to pad our quotas. We also kept quotas low and achievable, with accelerators. This enabled us to avoid the typical turnover in sales that you get in the Valley (15%-30% a year). With turnover low, sales reps happy and fulfilled, we weren’t vulnerable to salespeople doing our performance management for us, by leaving. I prefer to measure turnover of sales in their first two years on the job. Companies too often fool themselves that they’re turnover is, say, 17%, when it’s actually 50% or so of reps who leave in their first two years in the company.”
How did you manage your financials for The Street?
“We ran the company at breakeven non-GAAP but made absolutely sure to manage to cashflow positive throughout. For a company that is becoming a cloud business this is critical until you reach critical mass sufficient to generate ongoing operating profits.”
How did you keep founder-employees in the company?
“My experience was to rotate one or two of my ten direct reports each year in order to increase the ability of our management team to handle growth and increased complexity. Through the years, I’ve been recruiting exec positions throughout. The talent that gets a company to $10M is rarely the talent that gets it to $1B. I use a Skill/Will grid to assess how best to deploy key people. You may have all the skills, but without the will, it might make sense for you to something else, either in Splunk or another company.”
Ten Key Lessons for Leaders to Achieve Escape Velocity
Dear reader, I don’t know about you, but I counted at least ten leadership best practices related to not only fulfilling a management team’s performance commitments to its shareholders and employees, but also for becoming a relevant and powerful player in a fast-moving and challenging marketplace. Big data and IoT are two emerging categories that have been funded in the hundreds of millions of dollars (probably billions if you include big-company investments by IBM, Microsoft, Oracle, Cisco, GE, and many other tech-oriented corporations) during the past five or six years.
Here’s my list of key lessons, for you to compare with your own:
Sullivan trusted his instinct as the new leader to downsize the organization while doubling the price of the entry-level offering – a gutsy move when you consider that the board and founders wanted to go in the opposite direction.
Refused to fall into the “platform” trap – instead, his insistence on marketing to early customers what they believed at the time that they were buying, i.e. an IT systems management tool, not a generic platform.
Added a security application through acquisition, and hired domain experts in security in order to gain instant credibility with CISOs and their staffs. The first part of this new strategy isn’t so surprising, but the early and persistent investment in bringing on domain experts is something that I see companies routinely short-changing.
Showcased customers in seminars in every geographical market, to the point where Splunk holds in excess of one hundred such events each year today, with large attendance by customers and prospects. While many companies hold seminars and webinars throughout the year, Splunk has found a powerful way of leveraging their customers and consistently attracting large audiences, augmented by thorough and varied customer case studies and snapshots published and regularly refreshed on the company’s website.
Made intelligent use of freemium-based subscription pricing early on to enable technical users in enterprise customers as well as smaller organizations to try the product at minimum cost/risk. While not so extraordinary today, Splunk’s early use of this freemium-enabled pricing structure, which is based on usage metrics related to maximum daily aggregate volume of uncompressed data indexed (expressed in gigabytes per day), was quite unconventional seven or eight years ago. While at times controversial with customers who anxiously monitor their budgets and worry when they see the effect of usage spikes, the simplicity of this model has enabled it to stand up to market and customer scrutiny for several years now.
Established a repeatable company-designated sales process as practically a key condition of employment for sales reps to have a chance of being successful. All around the Valley and elsewhere in the tech industry, I still see way too many CEOs, sales leaders, and account executives winging it, working sales opportunities in their own individualistic, often opaque styles. Among the worst malpractices in this area are that (a) reps make crazy promises to customers in order to close the deal, while (b) keeping their “post-sale” colleagues (Professional Services and Customer Support in particular) in the dark regarding what the customer is expecting will be implemented, by when.
Made a habit of placing a greater emphasis on EQ rather than IQ in their sales teams. Our industry has a tendency to be besotted with technical or analytical “smarts”, while grossly under-estimating the importance of soft skills such as informed curiosity and genuine empathy. In my experience enterprise customers respond much more favorably to the latter than the former, which they often regard as table-stakes to get in the room.
Consistently provided conservative guidance to Wall Street, while also managing Splunk’s sales people, process, and quotas in such a way as to ensure that revenue achievement meets internal expectations and external guidance. Even more remarkable has been Splunk’s practice of making managers and sales reps accountable for meeting their “achievable” quotas with minimum extra cushioning for managers to compensate for shortfalls by some of their sales people. This practice depends on a culture of real accountability at all levels of the organization.
Eschewed the absurd but unfortunately commonplace practice of pressuring sales reps to learn how to sell the company’s products (too) quickly and perform to unrealistic quotas. This common malpractice merely acts as a predictor of high rates of turnover, for which sales management compensates by building in large amounts of quota coverage to their sales plan. Management teams in B2B tech companies almost never seem to compute the real costs of replacing 100% or more of their salesforce every three years or so in terms of hindering growth, increasing hiring and onboarding expenses, and operating the business sub-optimally with a demoralized sales force and support staff. Instead many seem to accept a 30%+ rate of annual turnover as normal.
Adopted a common-sense Skill/Will grid to assess not only the capability of senior executives reporting to Sullivan, but their reserve of continued will to grow new skills as the company grew. This is always a tricky topic for non-founder CEOs to navigate with their direct reports, and particularly with founders and early employees. Using a grid like the Skill/Will model enables leaders to have the “awkward conversations” be more candid and less awkward.
One can – and should – be skeptical about statements made by a business leader in a welcoming interview environment. That said, to my ears and eyes this was a hugely refreshing expression of – I hope you won’t mind me using the term again – deep common sense, combined with the courage to put it into practice in an often-challenging environment for growing from a relatively small footprint of $10m. or so in annual revenue to become a meaningful player among giants in the market.