This is a transcript of the podcast: Clearbit’s VP of Customer Success on turning economic headwinds into a growth opportunity.


Dee Reddy: Luke, you are very welcome to the show today. We’re delighted to feature you as part of our new ongoing series of scale. You’ve had a really impressive career to-date with experience that takes in hedge funds as well as customer success. To get us started, could you give us a quick rundown of how you came to be Clearbit’s Vice President of Customer Success?

Luke Diaz: Sure thing, Dee. Thank you so much for having me. To go back to the hedge fund days, I worked out of a long-short equity hedge fund in China where I had the experience of gallivanting around Beijing and Shanghai, taking care of our investors who had invested money with our hedge funds. That was about five, six years of my career. And then I got the temptation to join technology. I made my foray into technology as the first CSM [customer success manager] at a company called Optimizely. I’m currently the VP of Success at Clearbit, so it’s been a really fun ride.

Dee: And that first temptation, what was it that drew you to the industry?

Luke: Working in finance can be a bit stodgy. You’re wearing pinstripe suits. You’re meeting with these very wealthy individuals and it’s a little bit more formal. I think what attracted me to technology and Silicon Valley was the informality of it. And that the best idea always wins. It’s not so much who you know but what you know, and the truth for the vast majority wins and the best ideas in out. I loved that it was a little bit more informal and more of a fun culture.

Dee: There’s definitely very few, if any pinstripe suits in the tech world. They’re not a good look. But, regardless of that, is there any of your hedge fund experience that you still apply to your work today and what would that be?

Luke: It’s funny, you asked that because I thought about one of our customers a few weeks ago. This particular investor in our hedge fund was very particular and he had invested a sizable amount of money – call it was around four or five million with us. We would send them our monthly newsletters: “Hey, here’s how the hedge fund is going. Just want to keep you updated.” And he sent me a note and he’s like, “Luke, I really don’t need all this stuff from you. If you don’t mind, just send me a text with your monthly performance.” And I said, “Okay.”

The lesson for me there was customer service can be a differentiator, no matter what industry you’re in. Each person has unique support requests. They want to be talked to and communicated in a different way, so that was the lesson that’s always stuck with me long after the hedge fund years.

Dee: Gosh, that’s interesting, but I suppose it makes sense. Dealing with some of the richest people in the world would give you a really good sense of how to treat a customer and how they like to be treated. Tell me about your executive reading series before we move on because you must be constantly picking up new ways of thinking while you’re creating these blog posts.

Luke: Honestly, it was to keep me up to speed. I don’t know if you’ve ever had that experience where you read a book, you get really excited, and then someone asks you, “Yeah. So what was that book about?” There’s that awkward moment where you’re trying to recall even one sentence about what the book was about. I put myself on it and I said, “Look, if you’re going to read a book and spend… The average book takes five to eight hours to read, so I’m going to take some notes.” That turned into the DBT Ventures Executive Library, which now has approaching 4,000 followers and readers. It was really just to read two books a month and share the notes. When you share something it’s fun because it starts a conversation. That became a little side project to keep me learning, so I’m never in that awkward moment again.

Dee: That’s as good a reason as any. Aside from being essentially the Johnny Five of the tech world, in your day-to-day life, you’re Clearbit’s VP of Customer Success and, like for everyone in the year 2020, it’s been an interesting year. I think that’s probably fair to say of most companies, but it’s been particularly interesting for you. I’d love to hear how automation has been so important to your team, in particular, in weathering the last six months or so.

Luke: There’s no getting around it. This has been a fundamental shift in how people work and it’s both professionally challenging and it’s, on a personal level, also very challenging. Automation’s played a key role because the COVID-19 pandemic has really created two constraints. One is about focus. When economic headwinds exist, you have to really, really focus on your customer and double down there. And then the second constraint is efficiency. You have to do more with less. There’s less budget, there’s less resources. So there’s both a focus and a efficiency constraint.

Automation helped us with the latter, which is doing more with less. Specifically what that means is we have a certain segment of customers, which tends to do okay with Clearbit, but they don’t realize as much value as other segments. We leveraged automation to, instead of having a CSM own the relationship directly, we started to supplement and then eventually replace the communications with full automation. This even includes being able to quote DocuSign and close a renewal opportunity all without human involvement. You can imagine the efficiency gains there. Of course, we have escalation paths when that doesn’t work, but so far so good. Automation has really been our key to driving efficiencies in more scarce resourcing times.

Dee: That’s amazing. We’ll talk a little bit more about the importance of efficiency to a business later on, but let’s chat a bit now about the kind of symbiotic relationship that exists between data and automation. I mean, you really can’t have one without the other, so your business, Clearbit is so, so important to that intersection.

Luke: It’s been interesting. I think symbiotic is the perfect word. bad data can drive bad automation and excellent data can provide very thoughtful, very effective automation. I absolutely agree with the premise that they are symbiotic.

Clearbit is essentially a data provider so we stake our ground on the quality of the data we provide into other systems and then we eat our own cooking. We use Clearbit data to drive our business, whether that’s attracting new leads, whether that’s taking care of prospects, and then eventually, as customers, serving them using clean, relevant data.

Dee: Then thinking a little bit more about that intersection, where do you see content fitting in? Because that’s very much part of the relationship as well.

Luke: I’m very grateful to have a content team at Clearbit who produces very fun content, much like the content at Intercom. It’s fun, it’s engaging, it’s diverse media, whether it’s a podcast or a fun blog post or a video. We do work closely with the content team to ensure that message is getting out there. The customer success team can help give a further amplified voice to the great content that’s being produced.

We also have been doing quite a bit of content creation ourselves on the customer success team. We’ve been recording a lot of help videos, filling in the gaps to create lightweight bite-sized content, so that, as a Clearbit customer, you can learn about enrichment or reveal or whatever product simply by watching a three to five minute video, and help us create resources that drive that automation.

Dee: You’d mentioned to me previously that, in a certain sense, you’ve over-indexed on content on your backend because, at the end of the day, anyone can set up an automated email flow. But if that content isn’t tailored, you’re not going to get that important engagement.

Luke: Precisely, with the content and the structure of that content, whether it’s a blog post about a particular use case or a particular industry, we’ve seen a lot of interesting data on our end about what resonates with whom. So, in this ambiguous word “personalization,” which I’m not quite sure anyone can define precisely, showing the right thing to the right person at the right time is the Holy grail of marketing and service. If you have lots of unstructured content, it’s hard to get it in a position where it’s getting to the right person at the right time. You have to have the latent data attributes, the size of the company, the industry. Even the functional breakdown of your customer is very important because a head of marketing ops might not have the same interests as the head of sales. Respecting each person’s individual interests has been challenging and I’m hoping to spend more time on that versus the pure content creation.

Dee: That makes sense. Actually what you’re describing there – getting the right content to the right person at the right time – it really marries with what Intercom is trying to do with the customer support funnel. But, to stay on the automation topic for a little bit – I know you mentioned the feedback loop that you’ve created to quote and sign a contract – what other automation tools have you leveraged over the past six months to continue to drive the business forward?

Luke: I’d be happy to share some of those. One email tool we’re particularly fond of is customer.io, which drives a lot of our onboarding sequences. It’s such that, when you become a Clearbit customer, you obviously have to do certain things, same as Intercom. You have to set up the product, configure it appropriately, make sure all systems are a “go,” and then go live. We’ve powered an onboarding implementation flow, which has some business rules built in such that, if you haven’t done certain things that you need to do, it follows up on those certain tasks. That has been very helpful in making sure no balls get dropped in very busy times.

Onboarding and implementation is one use case. That’s the main one that comes to mind at the moment. We’re also doing more automated outreach on behalf of the CSM and the middle of the customer lifecycle. When we looked at our engagement patterns, we found that we were very active early on and very active at closer to the renewal. That made me kind of uncomfortable because there’s six months where value realization should be increasing. We saw a dip in CSM engagement and we were able to fill some of that gap with automated outbound emails from the CSM suggesting new use cases based on their industry and their product.

Dee: That’s brilliant. It’s a way of keeping people engaged throughout the cycle of their contract?

Luke: Exactly.

Dee: Then in terms of when that fails…I suppose the thing about the last six months that has been so challenging is that customers are under pressure themselves commercially. That email from a CSM isn’t necessarily going to save the business because churn has become a much bigger problem than it would have been this time last year. How has Clearbit addressed that issue of churn? Because I know you’ve done some automation internally that has helped with that.

Luke: That is a very hot topic. I imagine a lot of the listeners can relate to the dynamic we’re all facing where there’s a lot of fiscal pressure, a lot of economic pressure, and everyone’s being asked by their CFO or their COO to stack rank their software and cut if need be.

Some of the things we’ve done to combat those challenges have to do with operational improvements. We’ve also created some data science models to help us save or invest in the right set of accounts that could be at risk. I could, perhaps, dig into some of those two categories more specifically?

Dee: I would absolutely love that. Would love to hear about red accounts actually.

Luke: Sure. Red accounts is simply a weekly meeting. It’s about an hour and a half where every CSM has nominated a certain amount of their portfolio that needs some sort of help. Maybe they have a product issue and our engineering team needs to get on it. Maybe they’re struggling financially and they could benefit from a discount or some flexibility. Maybe they had a lot of turnover and we need to re-onboard the new team or whoever’s taking over, and we need to find out who owns Clearbit in a turnover scenario.

They bring their biggest churn challenges to this meeting. It’s not to blame anyone or point fingers. It doesn’t feel like a pipeline review. It feels like crowdsourced problem-solving. To make it fun we have a theme and we change our virtual backgrounds on Zoom to music or summer. It feels much less acerbic than the name implies.

Then you pose your challenge. You’re like, “I’m really blocked here. I haven’t heard from this customer in four email outreaches, what should I do?” And then the team chimes in and says, “Hey, have you thought about this? Have you thought about that? Have you leveraged LinkedIn Sales Navigator? Have you filed a support ticket? Maybe there’s a back door that could get this person’s attention.” And it becomes about creative problem solving. That’s been effective and it’s helped reduce churn by about 15 to 20% last quarter.

Dee: Wow. Then you assign businesses a churn probability score. Is that right?

Luke: Correct. We had the benefit of kicking off a data science project at the beginning of 2020. The goal was simple: help us figure out which customers have a high likelihood of churning, so that we can invest in those customers.

Without going too much into the data science weeds, we basically had a bakeoff between two machine learning models. One was Random Forest, and the other was XGBoost. If there’s any data science listeners, I imagine you’ve heard of XGBoost, which is the darling and recently won a prestigious data science award. It turns out that the Random Forest model was actually more predictive and more valuable than XGBoost for our use case. It took into account latent attributes like Alexa rank, Twitter followers, employees. It also took into account product usage and then CSM engagement. How much are we talking to the customer?

The output of all that was we have a likelihood to churn prediction score in Salesforce. I simply run the Python notebook and upload the data to Salesforce on a tactical level, and then that helps the CSMs prioritize. It also creates a fun dynamic where you have a quantitative yin to the qualitative yang, such that if a data science churn prediction score is above 35%, our threshold, the CSM has to convince the whole team that this account is not at-risk if they haven’t previously flagged it. It’s kind of a nice one-two punch between CSM insights and machine learning.

Dee: It’s interesting you say that because I was thinking that you are obviously going to find churn risks there that might not have occurred to the CSM based on instinct or their understanding of the account. This throws up some interesting data insights that you actually shine a light on then. In a way, that’s kind of the perfect marriage of how human instinct and knowledge can work alongside automation and data.

Luke: Yes. It’s been a fun quantitative safety net for our own intuition.

Dee: It’s actually working for you. It saves you quite a deal of revenue.

Luke: It has. The data science model has caught and helped us save a few hundred thousand dollars. It’s been meaningful to the business.

Dee: That’s a really innovative measure to undertake. I love how you’re doing it with the team, of having people step up and having to fight to say why they think a customer isn’t going to churn. It does add a bit more of that kind of human instinct, I think, which is the word that you used, and it’s the best one.

Look, these are all really, really innovative measures that you and your team in particular have implemented. How successful have they been for your organization in retaining ongoing profitability over the last while?

Luke: That’s an interesting question. Most tech companies are not profitable in a vacuum. They’re investing, or perhaps over-investing in growth at the expense of profitability. I’ll answer that question at my functional level where I can actually have an impact, which is customer retention cost. When you think about profitability in SaaS software, you’re typically looking at gross margins around 80%. Your biggest costs are typically your Amazon bill and customer success. That doesn’t mean that customer success is a cost center, but it is a significant portion of your cost to retain customers.

What I think an interesting exercise for any leader in customer success, or a CRO, to do is to actually take the entire cost of customer success – call it $2 million – take your head count, your all in P&L and then divide that into the revenue that you protect. Is the cost of customer success relative to revenue? Twenty to 25% like you typically see in pre-A-round or seed companies? Then, eventually you should asymptote, based on Salesforce data. Salesforce kind of sets the curve here that customer success is about 9% of their revenue.

So, one way to think about profitability is: what is the cost of customer success relative to the company’s overall revenue? Also what’s the ROI, like what does that get you? My team happens to own renewals and we generate upsell opportunities for sales. So, what I like to do is show the retention metrics, plus the revenue impact through upsell opportunities, and create an ROI calculation such that, in these very sparse times, in a pandemic, I can credibly say, “Hey, here’s what you’re investing in customer success. Here’s what it’s getting you.” That’s one way we think about both controlling our costs but also think about the ROI relative to profitability.

Dee: That makes perfect sense. It’s a really efficient way of thinking about it, which seems to be an ongoing theme in your thinking this year. I’d love to chat a little bit more about Clearbit ads, which you launched late last year. For anyone not familiar with it, you can give us a little bit of an explainer on the offering and how it differs now from when you first launched it not long before COVID unfolded?

Luke: Just to give a very brief history of Clearbit as a data provider, we actually started as purely APIs. Our founder and CEO Alex MacCaw was an early engineer at Stripe and he wanted to do for data what Stripe did for payments. We started as purely APIs, grew nicely, developed products that configured and integrated directly with Salesforce and Marketo and Pardot and so forth. Our product offering blossomed. Clearbit X is the third phase, you could say, of our product evolution where, for the first time, you can actually import your own data from any source into Clearbit X to enrich the targeting functionality and the targeting capabilities of Clearbit.

One example of this and where we’ve seen most of the early success is doing B2B targeting on Facebook. If you’re a marketer and you’re trying to get customers to join your webinar or sign up, a lot of that budget’s going to LinkedIn. We’ve heard from the market that it’s been challenging to get scale at a reasonable cost on LinkedIn. This isn’t a slight on LinkedIn. It’s just the economic challenges that prevail. We created a platform that integrates with Facebook and Instagram, and you can use Clearbit data to do B2B targeting on Facebook. That’s been one of the really interesting use cases. We call that Clearbit Ads or Advertising.

We have heard, in the light of COVID, the shift from growth at all costs or “I will need to grow, grow, grow” to “I need to grow more efficiently.” That’s been the biggest switch in our value prop, so instead of talking about total lead volume or total opportunity pipeline, we’re hearing more feedback like “I need to get the cost of my confirmed MQLs down” or “I need to have higher quality leads for the same or less amount.” That’s been definitely a switch we’ve seen from growth to efficiency in light of the pandemic headwinds.

Dee: That makes sense. It’s helping people harness the data versus to drive efficiency as opposed to driving growth at the moment – that’s the key message there.

Luke: That’s correct. We’ll see how it all shakes out as the world health situation unfolds. We’ll be watching that closely.

Dee: It’s interesting to me that you mentioned LinkedIn and Facebook. I wonder, has the blurring of boundaries between home and office affected the fact that there’s now more B2B businesses wanting to advertise on Facebook?

Luke: It’s been interesting in this new operating environment where there’s people taking board meetings in their bedroom. In no time in history has there been such a recognizable blurring of your personal and professional life. We have a hypothesis that we’ll continue to see that across the digital medium, where seeing a business ad on Facebook isn’t so strange when you’re taking a conference call from your house. In addition to LinkedIn, which is a very strong channel and avenue for a lot of advertisers, we actually believe that is going to expand. B2B targeting on Facebook has been, historically, very hard due to some data schema challenges. Thankfully we’ve been able to unlock some of those. We’re optimistic about the blurring of the home and office, expanding the receptivity, if you will, of an audience willing to click on a B2B ad on Facebook.

Dee: Have there been other ways you’ve noticed that the work from home reality has affected how people use your data in any way?

Luke: One of our products leverages IP intelligence to help who is visiting your website. Obviously in a work from home environment, IP targeting capabilities dropped off dramatically when everyone started working from home. We didn’t know what IP block you were on. That was a big challenge. A lot of people that do account-based marketing or targeting are facing this exact challenge because we’re on our home networks, not our corporate networks. That has taken a very special engineering effort to blend basically our people and our company graphs to basically reflect, “hey, this is a person who works at this company but they work from this home IP block.” You have to appreciate that targeting at least by IP block is much different than it was four months ago.

Dee: That’s really, really interesting. It’s not something that would have occurred to me, but the minute you say it, it makes sense. Staying on the topic then of working from home, I’d love to hear how your team has been getting on? I think, of all the roles in a typical tech org, customer success teams have had a difficult task over the past while, especially dealing with stressed out customers and having to do it in their own bedrooms or their own living rooms. How are you supporting them and maintaining sensitivity for the emotional component of their work?

Luke: It’s been hard. When we all started working from home there was a morale drop. A lot of people were trying to get their home offices set up and we saw morale drop very significantly across the board and that manifested in activities. We saw emails, calls pretty much drop 35% across the board. Thankfully, we work at Clearbit. I’m grateful to work for a company that values what we call “Conscious Leadership,” based on the fantastic book, which, if you haven’t read it, I highly recommend.

Our leadership team was very thoughtful and they said, “Hey, we’re going to bring some intentionality to this work from home challenge. We’re going to try and solve it together. It’s not going to be perfect, but we’re going to try and solve this together.” For example, they started rolling out these very interesting programs to help our Clearbit team improve our energy, our psyche, our morale. One example of that is thanks to our CMO, Matt Sornson. It’s a unique experience because it’s essentially a breathing exercise leveraging the Wim Hof Method.

Twice a week he does an hour long breathwork class where you’re doing Wim Hof Method. You’re on a Zoom call with 40, 50 of your coworkers and basically hyperventilating.

Dee: Who hasn’t done that at some point over the last six months?

Luke: You’ve got to do it because it breaks up the day. It is incredibly restorative because of the oxygen levels that infuse the blood. When I tried it last week – I’m not kidding – I held my breath for three and a half minutes after this experience. It was insane.

Dee: Wow.What’s the uptake for that, Luke? Do most people get involved?

Luke: We have just over a hundred employees or so, so there’s anywhere from 10 to 40 or 50. I’d say 10% to 40% of the company attends one of these a week, which is a really good showing in a remote-only environment. Just coming back to that theme of conscious leadership, the leadership team had everyone read this book. I think it’s in our DNA to say “Hey, we were having some really tough challenges. They’re not just professional. We’re all adjusting to this unique environment. We have to manage. We have to deal with the team, not the problem.” It’s been nice to see them investing in these really fun programs to keep things light and bright in a somewhat gloomy time in history.

Dee: That’s really nice. I think especially if a book like Conscious Leadership has really resonated with your leadership team, what better way to make that really clear to the people that work with them and for them then by sharing it.

Luke: Very much so.

Dee: Before we wrap up Luke, to actually stay on the topic of reading, I know you do your executive reading series that you mentioned earlier in our chat, and I’d love to know what your top recommendation for customer success leaders would be if you had to pick one book.

Luke: That’s a good question. You can see the full list at dbtventures.com in the library. There’s so many to choose from. In this environment, I’m going to go with Deep Work by Cal Newport.

I definitely wholeheartedly agree with Conscious Leadership, but since I already mentioned that I’m going to choose Deep Work by Cal Newport, which was first recommended to me by an engineer. The premise of that book is that, when you look at the data, we only are actually doing about two to three hours of work in a given day, given the amount of software and task-switching that the typical knowledge worker does. The premise behind Deep Work is you need to block two, three, even four hour blocks of time to get beyond the Slack messages, the email inbox, to actually get deeper into problems, to produce something like a podcast or something that’s more tangible and substantive. The only way to do that is through deep work, so I found that to be a very useful read at this time.

Dee: Well, we’ll definitely link to that book on the blog post accompanying this and anything else that you mentioned. Before we let you go, where can our listeners keep up with you and Clearbit?

Luke: I love doing coffee chats and LinkedIn is probably the easiest way. Just Luke R. Diaz on LinkedIn. I’m also on Twitter. I’d love to continue the conversation if anyone found this useful. Dee, you thank you so much for hosting me. It’s been a true pleasure.

Dee: Absolutely Luke. It’s been great chatting to you today. Thank you so much.