Very rarely do I come across a theme that I can be both dogmatic and contrarian about. More often than not, metrics, strategies, tactical responses are pretty uniform across venture-backed companies. For example, we can all agree that producing more revenue for every dollar burned just makes good, practical business sense. However, the one SKU that I adore, but everyone else hates, is managed services. Part of the reason I love it is because I’ve personally seen it work wonders during my time at CrowdStrike. Another part of the reason I love it is because I see how it could be beneficial both for my team and for other tangential teams. I’ll walk through why this worked so well at CrowdStrike, why I wish someone would operate this in the Modern Data Stack, and lastly, how you can implement it within your company should you feel so inclined to.
Love at First Sight
CrowdStrike (CRWD) started heavy in the Enterprise and moved down the ladder into the mid-market/SMB space. The thing about CRWD is that the product is highly technical, and the meaty product — Endpoint Detection and Response — is really best served for companies with an in-house SOC. While CRWD could have relied on its Next-Gen Antivirus to carry the load downstream, the reality is it would not have taken them very far. Not only is the antivirus space very crowded, it also comes at a lower price point and doesn’t accomplish the task at hand — protecting highly vulnerable customers from any threat. (Actually the mantra used to be ‘We don’t have a mission statement; we stop bad guys, period). Yet, it was tough to push the meaty product downstream since many companies struggle to hire cybersecurity talent; indeed, the country as a whole is facing a dearth of cybersecurity talent.
So what was the solution? EPP Complete. We were fortunate in that we had an in-house consulting arm (think Deloitte, E&Y, etc.), so we had an understanding of what kind of folks would work best in this virtual SOC. With EPP Complete, CrowdStrike was able to offer all their products + a virtual SOC on top of the product offering to deploy to the middle market. This became the stone that hit multiple birds, in the following ways:
We could now deploy all our products into the middle market space; adding data points from companies who might be susceptible to anomalous behavior (e.g. bad guys)
We could use the information from the endpoints to strengthen our product offering
If any bugs existed they could be triaged and found quickly by using the product ourselves before our DIY users found out
Companies no longer needed to pay heavily for cybersecurity talent that may not have wanted to work there in the first place (if you’re in high demand, why work for a no-name shop stopping phishing attempts?)
We could charge a higher price per package (since the customers are getting a white glove service) while also saving the customer money and benefitting from economies of scale
The results were really quite impressive. At IPO, CRWD disclosed in the S-1 that their professional services arm generated an additional $2.97 ARR for every $1 in Revenue generated in Services. That’s on top of the fact that the services themselves operated at a 44% gross margin.
To conclude on this piece, the Managed Services and Professional Services arms really help companies like CrowdStrike deploy more product to their customers, generate economies of scale while passing off that economic profit to customers, and supercharge revenue growth. Of course, Newton’s 3rd law can’t be ignored (or Thomas Jefferson in Hamilton). Market share will eventually erode from third parties, but that’s OK, because the customer is left better off staying in the CRWD ecosystem.
Boy I’d Love It If Someone in the MDS Did This.
The modern data stack needs this. The interesting thing about the MDS is that each unit is SO powerful as a stand-alone product, yet completely useless without the other parts.
Data sources -> EL -> Data Warehouse -> Transformation -> Data Warehouse -> Reverse ELT -> Data Sources
Why would someone in the MDS benefit so much from offering managed services? Because a majority of companies are nowhere near analyzing and exploiting all their data and a majority of companies have accepted having parts of the org be completely data-illiterate. I’ll use Human Resources/People Operations as an example here. The excuse I usually get for excluding HR data from the Data Warehouse is something along the lines of “It has really sensitive data that no one needs to see”. The consequences? Most teams can’t tell you where the inefficiencies lie in HR, so we’re left with a growing amount of organizations that lack transparency on important topics like diversity and inclusion.
Who would benefit the most? Pretty much every step of the MDS. A cloud data warehouse, like Snowflake, would grow compute/storage revenue by offering managed services. A cloud data replication tool, like Stitch, would grow consumption revenue because the rows produced would increase. Lastly, a BI tool like Looker that charges by seats would need to increase their seat count at each customer. These are all direct economic benefits, but what I won’t dive too deep on, because they’re intangible benefits, are the benefits of being the quarterback of the entire deployment, the lifelong consultant for the end customer, and the beneficiary of doubt.
What would it look like? I think if someone like DBT offered this, I would imagine they would charge something like $100K/Year for a fully managed offering, maybe up to 20 data sources. They would then go ahead and white label these powerful sources and have an in-house data analyst act as a consultant to the companies to find out everything they’re missing out on. In fact, $100K might even be on the low end. A good data analyst team should be able to find lots of uncovered gems — how to best spend marketing dollars, which leads should go where, which territories make sense for GTM expansion, how to optimize pricing etc.
How to Do This
Everyone is going to tell you not to do this, and maybe you shouldn’t. Being a trusted advisor to your customer is helluva lot harder than it looks. It sure is easier to sell shelfware and move on with life. Bonus points if you spend millions on a customer success team that operates like a group of chickens with their heads cut off. Extra bonus points if you layer a finance team on top of that that says Customer Success has a low Net Retention and Poor ROI. But how often do you hear about a fantastic software company with a top-shelf managed services offering going under? Fuhgetuhboutit.
Develop great partnerships — find yourself in a sea of tech partners, each of whom doesn’t feel like you’re introducing on their territory
Develop a great standalone product — your product needs to be essential; if not, you’re just a consultancy
Keep Your Friends Close and Enemies Closer — Don’t turn off other consultants, not every deal will fit under the umbrella of managed services. Throw these leads to the Platinum service partners
Price correctly — Don’t be a hog; make sure your passing off the economies of scale to your customers
Treat your customer the way you would want to be treated, and then a little better — you want to be a trusted advisor, give them the time they need to feel a sigh of relief