Jason Cohen is the founder of WP Engine, who went from $0 in 2010 to $132m in revenue today with 90,000 customers (and a $250m investment from Silverlake last year).
Jason started WP Engine in 2010 which he bootstrapped for two years, but then raised money and rapidly grew to 475 employees globally, serving 90,000 customers and still growing at a healthy clip.
Prior, he bootstrapped Smart Bear for seven years, built it to millions in profit, and sold it in 2007. It was sold again in 2017 for $450m. Before that, co-founder of IT Watchdogs, which sold in 2004 and is still producing its products under its parent company Geist.
Also a mentor at Capital Factory (like TechStars or Y-Combinator in Austin) and an investor in a few companies including StackOverflow, SpareFoot, Cratejoy, VolunteerSpot, SwimTopia, and Pingboard.
Follow Jason on:
WP Engine recently took $250m investment from Silverlake, tell us about the difference between dealing with traditional venture funds vs. big PE like Silverlake?
We took that almost 2 years ago, which is good because that provides some post-raise perspective. PE is not venture. Venture expects many bets to go to zero or <1x, so the remainder need to be as large as possible, even 10x is just OK and often 25-40x is needed to return the fund.
PE (or other late-stage) invest in known quantities. Enough scale to be predictable. Almost nothing goes to zero. Big success is 3x in 3 years or even 4x in 5 years.
Different PEs are different. Some buy and flip or tear apart or reorganize. SilverLake buys and invests in innovation. So they have a long time horizon, they care about lots of happy customers more than short-term metrics, even important ones like GPM. Better to have great metrics on e.g. retention and NPS and solve costs later. So they’ve been a great partner in investing in innovation and customer experience.
Would they fund more burn as long as you have 120% net rev retention, low churn, and other economics generally healthy?
Yes, and so would any investor, probably at any stage! Just mathematically, that works. As you said, if e.g. GPM is low and/or falling, or N churn is high and/or growing, or CAC is high and/or growing, then maybe not, althouh even then, probaly worth placing more bets to see if that can be fixed. Especially if you have 120% at scale (by which I mean >$100M, or even probably >$50m)
Thanks for making the time! How has your experience been with the Austin tech scene?
Recruiting is tough — Austin has 2.7% unemployment, lower than Bay Area. Housing prices are skyrocketing. Infrastruture is over-full as Austin has been %-wise one of the fastest-growing cities in the country for many years now. If you have more specific questions about raising money or company types or support for certain verticals, LMK!
I talked to a SV VC last week and he mentioned that as we are in Dallas no SV VC would invest in us. Do you agree with that?
If you’re early-stage, then that’s accurate. Series B maybe, Series C you can raise from anywhere. But that’s not just true of SV. Few early-stage VC wants to invest outside of where they physically are. You stay very close to companies early-stage, and helping them also requires a network (e.g. helping with key hires), so it makes sense. Unfortunately it makes it difficult for new metro areas to get started.
Do you invest in early stage companies?
I don’t right now because I’m too caught up with WPE. Haven’t made an investment since 2014.
Hey Jason. Thanks so much for joining. My company (Seahawk) is a GoDaddy and WP Engine partner - we prefer WP Engine for most of our woocommerce sites as other Managed WP providers can get bogged down. We are familiar with the Agency Directory for WP Engine partners. Does WP Engine plan to offer professional services like WordPress design, development, tweaks in-house? If not, how do you guys help customers seeking these services.
Great to hear! No we will never offer hourly or project-based services. The reason is, that would compete with our agency partners. We want to help grow business for our agencies, not compete with them. We are constantly referring customers to agencies; the agency partner program is part of how that works as you know. If you’re not seeing a lot of that, you should talk to your rep to find out how to get plugged in! That might involve e.g. being more specific on what you’re excellent in, so we know how to route requests.
How often do you guys look at your pricing? Traditionally how often does it change?
We’ve changed prices once in 2012 and again in 2018, so “very rarely.” The earlier you are, the easier to change or test pricing. Once you have e.g. 100,000 customers, any change is massively impactful, comes with communication, is judged by others, competitors launch marketing campaigns about it, and so on, so you have to be careful and not be constantly doing that.
The key is to hear from customers. The first time, it was because customers said things like “I can get my client to move from $5/mo shared hosting to, say, $30, but I can’t get them to $50.” Of course that’s physcology more than anything, but we heard it a lot. So we reduced our lowest plan from $49 to $29.
We also heard “Look I have a big site I’m happy to pay $99/mo for, but I have this site from my mom that gets zero traffic. Why can’t I just do that too?” So we added “up to 10 sites” for the $99/mo plan.
When you’re reacting to that sort of thing, it’s good. If you’re just trying to squeeze more out, that only works if you hve a lot of competitve power (in the Michael Porter sense). Make sure you have that first.
Some light reading on WP Engine while Jason is busy typing https://wpengine.com/blog/wp-engine-announces-132m-annual-recurring-revenue/
Hey Jason, thanks for being here. What were your best working non-paid marketing channels when you were scaling?
Twitter. We’re lucky though — our audience is WordPress site makers, and they talk! They blog, they go to events, etc.. Other markets might not be like that.
That said, nearly all our scaling was paid. I’m actually not a big believer in unpaid. I think even when that works well, you spend a ton of time there (which is also money, often more than paid, because it’s people), and sometimes you end up being dependent on just one person who is awesome at it, and then they leave and you’re in trouble, unlike more systemtic campaigns.
Pre-scale, nothing I just said matters. Just get the people in! But scaling means being systematic.
Big WPEngine fan here, thanks for your time and for creating this awesome solution. Since you founded WP Engine, is there any specific stage that was more challenging than the others?
Thanks so much All of them are of course challenging for different reasons. I feel the bigger, the harder, however. The whole thing of “I’ll hire [leader] and then it will be easier” is incorrect. It is true that you have to do that, so you can be more excellent and so you can tackle the harder and bigger challenges, and be strategic and so on. But it’s not true that the problems get easier.
They’re harder because they’re more complex. Finding a new product offering is harder because if it’s not $10M ARR or more in potential, it’s not worth doing. But very few ideas are like that! This isn’t a problem when you’re small.
Or people. You might have a group of 60 people who aren’t being strategic or too slow or etc etc, and solving that is very complicated. Not like a single low-performer who you can manage out.
Even communicating thing slike strategy or changes is very hard. We have 900 people; no one hears everything or hears it the same way. 5 people in a room can be synchronized easily.
Guess I’ll jump in here - How do you think of competition from all of the upstarts who won’t be using WP in the next 5-10 years (Gatsby, etc)
We have customers using Gatsby so….
Two answers. One, that same question has been valid for the entire 10 years we’ve been in business, and since then WP has gone from 12% to 34% of the internet. That includes the rise of mobile phones, social media, Wix, Squarespace, and so on. So one answer is, the WP community is robust. Even if that slows (doesn’t it have to slow!?!), it’s massive so it’s not an issue.
Second: Most of the software we build is outside of WP. WP is the CMS for our customers, but the dev tools are from us (and yes we’re getting into headless there), the intelligence tools, some things we’re still working on, they’re not WP. That has to be the long-term view, so that WP becomes a component like Linux or MySQL and not the whole thing.
Hey @Jason Cohen, how does your paid marketing channel look like? Where do most of the money go to, Facebook ads? Google search ads? Podcasts? Or promoting on micro niche communities?
We spend about $2M/mo on paid channels, so the answer is “many things.” We maxed out channels like Google early on.
People often think there’s “always more” there, but generally it’s the opposite — it gets consistently more expensive and lower quality every year.
We constantly look for new channels. Affiliates is an itneresting one, because it’s difficult, it’s a zero-sum game, and we’ve variously been good and bad at doing it.
I have not seen you guys yet on TV? Any plans? how do you decide on picking up a marketing channel to invest? I would love to know your challenges around marketing.
TV isn’t how agencies decide which WordPress platform to use. That said, we haven’t tried, so I suppose it’s possible. But, e.g. Adobe doesn’t advertise Experience Manager on TV either.
What we do is go wherever our customers go. Events, online, etc.. Where do agencies go to conduct business, or learn about new tech, etc.. That’s where we go. GoDaddy is reaching millions of people who make the decision themselves. None of the enterprise CMSs advertise on TV that I’m aware of, because that’s not how multi-million dollar decisions are made.
What were the primary drivers behind the decision to acquire Flywheel?
FW focusses on the freelancer and small agency. They have great tools for that, as well as the brand and voice to match, as the founders are that themselves.
WPE focusses on the mid-market and enterprise, although we have 10s of 1000s of smaller developers as customers too (because we’re the market leader in managed WP).
So first: It was a “puzzle piece match” in terms of market coverage from a brand and prodcut perspective. On top of that, a great culture-fit, especially in being customer-centric and in how we treat our employees.
Additionally, “Local” is the leading local dev tool for WP, we had seen early traction with our own tool, and that was a way to integrate and be better together.
Jason what are your favorite tools/websites/apps that you use as inputs to decide on strategy?
My favorite book on how to build a strategy is Good Strategy Bad Strategy. It explains why a lot of things people label as “strategy” is actually not, and is very clear how to build one that is, and why. It probably could be 1/3 the length, but it’s good.
Tools – None, just the usual docs and presos and such. Inputs – I have a specific thing I do here:
1 Gather data. Often you don’t have the data you want, but gather what you have about your own customer behavior and the market.
2 Debate your “convictions.” These are simple statements that you believe, whether you have data or not. These will be treated as axioms for purposes of strategy. Everything from what’s important in the market and product and customers and pricing and where things are going and what they are now and so on.
3 Try to validate convictions even if qualitative (e.g. customer interviews).
4 These being key inputs that help guide decisions and ID topic clusters that become your pillars.
So, first there’s the system described in that book which is:
1 Define the problem/oppy (they go into detail on what “define” means, so you’re being crisp, really digging into the truth and heart of the matter, etc)
2 Explain a few main themes/pillars that will allow you to attack the problem (3-5). These are organizing principles that lead to focus.
3 Create specific actions (e.g. like a meta-backlog) fro those themes. This is where you have detailed projects/products/department-level or multi-department programs that execute against those pillars.
So some of the above is part of getting to (1) and figuring out what qualifies as smart for (2) and (3).
OK so next another trick is to use something called “tensions.” The idea here is that you need more things that help you make strong and difficult decisions.
Because you’ll end up with lots of ideas where you say “well X is better than Y along metric Z, but it’s the opposite for metric W” and you go around and around and can’t really prioritize or decide what to say “no” to. So tensions help break that cycle. A “tension” is just like the Agile Manifesto, where you acknowledge “both of these things are valuable and good, but one of them is going to win when there’s a conflict.” Like “People > Process.” Here’s a tangible example for StudioPress, a brand we acquired last year for WPE, when we built our post-acquistiion strategy: “guided upgrades” > “backward compatibility”
A bad way to write this would have been “new secure stuff” > “people running old code” because then it’s not in conflict. And doesn’t honor the fact that backward compatibility is a virtue (with the negative side-effect of old stuff in the field).
Here, we’re honoring the benefit and drawback of both things, and explaining we’re going to pick just one of those things, but how we’re giving up something on purpose (back-compat) and how we’re going to make our choice the best-possible for customers (don’t just force upgrades, intentionally make it easier).
We made someting like 2 dozen tensions for various observations and pillars with the SP strategy.
The discussions around tensions are awesome too! They’re constructive because you’re not saying “one of these is bad.”
And because each one is individual, the discussion can actually happen without too much distraction. Real projects have multiple dimensions and that makes them hard to compare.
One you have tensions, some decisions become clear, fast, without much discussion. That’s their power. Plus they’re further documentation fo “what we believe.”
Thanks for doing this AMA Jason. How does the CTO role evolve from writing the first line of code & being bootstrapped to helping steer the ship at 900 people and $100m rev?
Well of course it goes to not writing code. However, if you never get your hands dirty, you don’t know anything and you can help guide good decisions. But knowing the business becomes more important than knowing the code. At 500 people, or even 100, if you don[t have people you trust to know the code and architecture etc., you haven’t hired properly.
ThereforeI stick to “hackathon” things WRT writing code. Meaning: POCs that aren’t products and aren’t supposed to be. Keeps me sharp.
Reason why it can’t be related to real products: Then I’m encroaching on the team. The implication is they have to use it.
Consider this: The fun part of engineering is inventing the stuff. 95% of the work is doing all the details of implementation, testing, fixing bugs, the 10% that is “the other 90% of the work,” etc.. The really fun part is you get to invent how to do somethign and then see yuour ideas come to life.
If the CTO steals that part, that’s a travesty for the team. You’ve stolen the fun part. Not OK!
Hey Jason. Thank you for doing this (thanks Nathan!). I’m a young founder from South Africa and have been following a lot of your talks the past few weeks and so I’m pleasantly surprised. I have a couple of questions which I would appreciate your input on:
1. As a bootstrapped founder how should one evaluate various startup opportunities and which to act on? I’m currently in the situation where I’m really passionate about the one startup, but working on another one which I believe will help me to a sustainable bootstrap level faster, but I’m not as passionate about it.
2. On marketplaces, from your experience why did SpareFoot not go all in by allowing people to book online versus calling a number? 3. From your experience, what is average monthly churn and best-in-class churn for B2B and B2C.
1 Never do X “so that I can in fact do Y.” Bad reason! It woudl be easier to have a regular day job with fixed hours, reliable pay, and doesn’t drain your energy, and do X on the side. Rather than do Y which takes 100% of your time and money is difficult, when you want to do X.
2 SF couldn’t do that because the booking systems for shared storage are archaic. Not connected to the internet. They did build a connector eventually that could do soemthing like that. Eventually they decided to merge with a company that makes software like that, so they could go end-to-end. So it took 11 years but they did that to some extent!
3 Really varies, e.g. Shopify’s churn is horrifically high but they’re a great company. If you force me to say, I’d say 2%/mo for SMB, 3%/mo for consumer, and <1%/mo for mid or larger.
What changed after SilverLake came in?
They are very long-term focussed, and they have deep pockets (i.e. “there’s more where that came from.“) Therefore, the focus is on product innvoation and customer delight.
Innovation because that’s how you remain relevant and separate from competitors. Delight because that’s how you get >100% net annual retention which is not only evidence that you’re doing valuable things for customers (unless it’s because you’re a monopoly) and because that’s the key to growth at scale (mathematically).
Also board meetings are easier! (Just one investor and independents, instead of multiple investors.)