Building Marketing Channels Slowly to Achieve Massive Scale with Datadog CMO Alex Rosemblat

Name: Building Marketing Channels Slowly to Achieve Massive Scale with Datadog CMO Alex Rosemblat
Uploaded: 2022-01-13T16:00:11+00:00
Duration: 22 min 49 s
Channel: Jason M. Lemkin
Description: In this talk, Datadog CMO Alex Rosemblat shares the framework he used to grow Datadog’s marketing from zero to a multi‑channel, revenue‑generating…

13 Jan 2022 · 23 min

Watch on YouTube

22:49

In this talk, Datadog CMO Alex Rosemblat shares the framework he used to grow Datadog’s marketing from zero to a multi‑channel, revenue‑generating engine. He argues that marketing channels should be treated like products: start with a hypothesis‑driven pilot, iterate methodically, collect directional data, and only after mastering a channel should you scale aggressively. He stresses the need for singular focus—one channel, one campaign, one team at a time—while explicitly declaring what will not be done to manage internal expectations. He walks through the trade‑show channel (AWS re:Invent) as a concrete example of a channel that was piloted, documented in a playbook, scaled, and eventually handed off. He also covers practical tactics such as setting minimum success criteria, negotiating with vendors, using negative keywords and geo‑targeting for quality, instrumenting reliable metrics, and recognizing the hiring inflection point (the “10th hire”) where team structure must be re‑engineered for scale. The overall message is that disciplined, data‑backed pilots, clear scope, and continuous optimization enable marketers to hit the ceiling of a channel and then move on to the next, ultimately driving sustainable revenue growth.

Chapters

My name is Alex Rosenblatt. I'm the CMO at Datadog, and Datadog, for those that don't know, is a monitoring and security company for modern IT infrastructure, usually cloud infrastructure, but we'll take anything that's, uh, that's complex. Um, I've been at Datadog for 8 and a half years now. At Datadog, I was the first go-to-market employee, so it was a bunch of engineers when I joined. I'm actually a former engineer, so that helped in understanding their language, but there was no marketing, there was no dedicated sales when I started. We started from scratch and it was a blank slate, and something that we very quickly found out was that building a marketing channel takes a lot of work and that it's something that really needs to be done slowly, steadily, methodically in order to be achieved well. And once you figure out a channel, then all of a sudden you can scale, and scaling becomes a lot easier. So on this, a good way to think about everything having to do with marketing channels is that it's a lot like a product. And maybe some of you have built product. I'm sure that many of you have interacted with product teams. Something that I think is emblematic of working with a product team is It's never done. I mean, it takes you forever to get to V1, and finally when you get to that first version, what's the next incremental version? 1.1, then 1.2, then sometimes you get 1.2.1, and it just never seems to be ready to get to that version 2 or version 3 that's a more mature, more stable, more robust product, which is, at least for a milestone, what everybody had intended to get to. And I don't think that a marketing channel is any different. It takes a lot of iteration, a lot of cycles, and a whole ton of feedback to get the marketing channel right so you can figure out everything that you need to do to scale and hit a ceiling. Now, a difficulty with building out marketing channels, especially in the amount of time that's needed, is that the feedback can't come from a peer or from a very small group of customers. It's gotta come from random people that could be prospective customers out in the market, and it takes a very, very long time for some kinds of campaigns, for them to take a look and then vote with their feet, vote with their mouse clicks, vote with their eyeballs to go forward and check you out more, or to say, "Eh, not really interested," or "Doesn't really fit my need." So you need tons of those cycles and the tweaks that you make from the feedback that you get to try to get everything right to build a marketing channel like a product to the point where you can scale it out and hit a ceiling. Now, if you're a marketing leader here in the room or you're working with a marketing leader, where things get a little bit tricky to get this singular focus is that the expectations that people have of marketing are huge. From PR to advertising, analyst relations, blogs, SEO, the list of things that everybody thinks marketing is gonna do is quite, quite expansive. And management thinks that marketing's gonna do one thing. Sales thinks that marketing is gonna do another thing, product something different, even down to an individual employee that wants to show the website off to their friends and family for the company that they're working at, and they've got expectations of what you're gonna be doing with that website as well. So with all of these expectations and everything that everybody thinks that marketing should be doing, I think that a key issue with a lot of marketing leaders or marketers inside of companies is that they very, very quickly get spread spread thin. They're doing a pinch of blogging, they're doing a pinch of analyst relations, a pinch of press outreach, a little bit of product marketing here and there, and when they're doing a little bit of everything, it becomes nearly impossible to become amazing at one single thing. Oops. So amazing that you want—
sorry, can you go back a slide?
Thank you. So amazing that you figure out what it takes to basically throw the kitchen sink at a marketing channel to hit the ceiling that might be there. A key question to ask here while people are doing a little bit of this, a little bit of that is, is any of this work actually turning into revenue? Because at the end of the day, the marketing function is very distinct. The marketer is trying to reach random people that could be prospective customers floating out in the market, trying to engage them, bring them back to some property of the company. Whether it's a physical trade show booth like the ones that we have right out here, or a landing page, a blog, something to show engagement, something to be brought in. And when you take a look, many of the activities that marketers oftentimes get sucked into or think are part of their scope don't really contribute to the bottom line. A lot of them are supporting activities that are there to make smoother those revenue-generating activities, and at certain stages of growth, if you don't do much with them, they're pretty much not gonna contribute to revenue. I think a really good example here, and of course for some companies it might be very important, but something like press, PR, press relations. You can do a super good job, you can come up with a great press release, get customer quotes, get customers to speak on your behalf to a reporter, engage a reporter, have your briefing with them, and then some news for a third-party company, maybe Apple, maybe Google, comes out at the exact same time that you had sent your pitch out and the writer says, "Oh, that's not as important as this other news," writes that up and all this work that you did literally goes for nothing. And that was time that could have been better spent definitely focusing on something that would bring in new customers and new revenue to the company. Again, if that is the goal for that company for marketing, which I think for most companies it is. So what we have found in 8.5 years at Datadog is that moving slowly, moving steadily, Moving methodically, moving with a lot of introspection, taking a look at the data for a pilot experiment is the way to scale marketing. And the thing is that similar to the hair, once you start to get these little insights, you found a little trick that works here, a little trick that works there, and then you start to use them all together, at some point you just start to make these huge leaps because you figure out what works for a marketing channel and before you know it, you're at a point where you can't grow that channel anymore because you have hit the ceiling. There are no other people available in the market that could see this particular ad or trade show booth or whatever it is to bring into your company property, and then it's time to turn around and start focusing on the next channel. So this all sounds good and we talked about how marketing means a lot of things to a lot of people and people are getting pulled pull in a bunch of different directions, in order to be able to really hunker down and do one thing and become amazing at it, it's important to first narrow your focus and then set goals realistically. So from what I have seen, an individual or a team, and here I'm defining a team as a set of people, it might be employees, it might be some employees plus contractors plus outside agencies, can only really become fully dedicated in one thing at a time. That means that they're gonna become experts at that one thing. If you work backwards from there, that means that you have to pick one marketing channel at a time for that team and likely only one single campaign or one single type of campaign, and that's all you get. If you're gonna become amazing and really understand the ins and outs of this one individual channel, you can only focus on one major thing at a time until— You should set some sort of criteria where you hit a milestone that you know that you have gained enough mastery in this particular kind of campaign or marketing channel that you're good, and then you can shift your focus to keeping that motion going and moving over to do another thing and start from the beginning again. One thing, one campaign, one team, singular focus, and working on that particular goal. At Datadog, The first thing that we did, that one thing, was sponsor trade shows, booths just like the ones that we have outside of here. I was lucky, again I mentioned I was the first go-to-market employee. When I joined, there was a calendar of events that fit our target demographic that were coming up, so that was the one thing that we did, and just getting really good at doing trade shows in a small company took literally everything that we had, and we really didn't focus on anything else for about a year and a half. But 18 months later, we were amazing at trade shows. We had written down a playbook. We could pass it off to somebody else to keep that going, and we turned around and started focusing on the next thing. This also means, because you're gonna be singularly focused, that you have to explicitly declare the things that you're not going to be doing. Now, this is gonna probably run afoul of a lot of expectations, of other people at the company that expect marketing to be doing one thing. And realistically, there's gonna be website changes that need to be made, there's gonna be the occasional blog and press release that needs to be released because the product team came up with something new, but if you're able to say, we're not gonna go heavy on press relations, we're not gonna go heavy on analyst relations, we're not gonna go heavy on content, we're not gonna go heavy on advertising, whatever it's going to be, if you explicitly state you're not gonna be doing these things, it becomes common knowledge that this is not a thing that the marketing team is working on at that exact moment. And very importantly, it's gonna be at first quality versus quantity. Anyone here that's run like a Twitter campaign or Google AdWords campaign, it's super easy if you set things pretty broadly to be flooded with attention, clicks, leads, what have you. You get a ton of them if you want, but are those the people that you really want? Are they really your target demographic? And that's where negative keywords and geolocation or geo-restrictions start to come into effect. You wanna hit that quality level and get a vein of exactly the kind of people that you wanna reach out to before you start plowing tons of time, money, and effort into scaling a channel. Now as I mentioned before, product sales, other people at the company are gonna have a lot of expectations, but the people that you have to get on your side are the management team and oftentimes the board of the company. So the management team is gonna be expecting a certain amount of things to happen from the marketing function concurrently. Declaring your scope and that it's gonna be super limited aside from just keeping the lights on in other areas, I think is hugely helpful.
Sorry, can you go back a slide?
Well, is important to making sure that management buys in and then also advocates on behalf of the board. An important way to get this message across is to establish the hard data that you're gonna need to evaluate whether you're hitting your goals. And in this case, if you can align the goals to the ultimate, the main business objective or the ultimate goal of the company, that usually goes a long way to supporting a singular focus on one marketing channel because the trade-off that you're making from not doing many things to doing one is that you're going to increase the reliability that you're gonna hit that goal. If you need 500 customers by the end of the year and that's gonna translate back to X thousand leads, you're gonna more reliably be able to hit those X thousand leads by doing one thing and becoming great at it rather than by spreading your bets and doing a lot of things maybe halfway, which goes back to data collection instrumentation. If anybody here is currently engaged in marketing and you don't feel like you have good data to work with, that should probably be the first thing to focus on so that you're able to get a data-driven case made to the management and board of your company to say, "We're gonna do this one thing in isolation until we become great at it." A note on marketing data, especially if anybody here in the room is working with marketing people, is that it's oftentimes imprecise. I've had the experience where you have multiple systems all trying to instrument the same metric and one gives you the number 10, the other gives you the number 13, and one gives you the number 22. They have different definitions of what the right metric is, of how they're collecting that metric, and while individual data points or individual data sources might be imprecise, the important thing is that they should be directionally correct. So if one data source goes from 10 to 40 or 10 to 200, you know for a fact that that marketing channel is growing and that even though the exact number might not be right, the direction and the scale of the direction for that specific data source should be something that you can trust more, and you know whether things are going in the right direction. Ultimately also, if you know what the case is that needs to be made back to the board and management of a company, you can work backwards to figure out the exact pieces of data that need to be collected, and again, with a laser focus, do a good job of instrumenting those few pieces of data to support the case that you're making about whether the marketing channel that you're scaling is indeed going in the right direction. So to get to all this, this is, I think, pretty theoretical. You're going to only be able to really focus on one thing at a time, so you want to make sure that you focus on the right thing to spend all that time on. And the way that we've done this at Datadog is that we work on hypothesis-driven approaches to different campaigns that we're gonna invest in, or even if it's a small monetary investment, it's gonna be a very large investment of a scarce amount of hours from the team that needs to get all the pieces together to make a test work. So someone's got an idea for a campaign, fantastic. Has the thought been put on paper of how this campaign is gonna work? Specifically, what's the mechanism? What's gonna be the time limit to assess success? Are there budget limits? Do you have enough expertise on the team or from the vendor to make this campaign work as designed? Importantly, what is the minimum result? I think that's a really important concept. The minimum result that you need to see to know that this specific pilot or this specific test is actually working the way that it has to. If it falls short of that minimum result, the thing might not be—
Sorry, can you go back a slide?
The thing might not be working at all. You might want to actually cut bait quickly and move on to another idea that's more likely to work out and become a beachhead for you as a company to start investing in and growing a marketing channel. Importantly, if you're working with an external team or an external vendor, you really want to make sure that they're going to be the right people to take this forward, and I can tell tons of stories of sketchy vendors that have done weird stuff that you were never expecting and you have to ask them to pull back really quick. Two other, I think, really important points are, once you put everything on paper and you're feeling good about it, it's very easy to become blinded to any weak points or any holes in your plan. Finding a peer that knows marketing to play devil's advocate and try to point out holes in your hypothesis is super helpful. Then lastly and extremely importantly, have you instrumented the data to make the assessment of the campaign success accurately? Do you know that you're getting the right leads collected from this source if it's a lead generation campaign that you will be able to very cleanly attribute back to that campaign to be very confident that this thing did indeed work? And once you have this, this chart has probably been the methodology that's gotten us to very, very scaled out areas in a lot of campaigns to the point where we've hit tons of ceilings and we can't expend any more time, money, or effort to grow that particular campaign anymore at Datadog. The number one thing is the pilot. Now a lot of vendors usually have a minimum amount that they want you to spend. Oftentimes those minimums are negotiable, especially if you're talking to them in the last—
sorry, can you go back?
In the last couple of weeks before—
can you go back to the previous slide? Thank you.
If you're going back to the vendor, working with them the last couple of weeks before a quarter, will usually get you the discount to the minimal amount of spend or the minimal amount of time that you need to run something in order to assess whether it's working or not. I think that that's another big concept. Time is scarce. Budget is scarce. You want to have the minimum amount that you need in order to see whether something is good and warrants more investment or whether you should get rid of it. Once the pilot proves out, you want to usually repeat and enlarge that pilot, make sure that there wasn't some sort of seasonal fluke, or that you just didn't get lucky, and this might actually be repeated a couple of times over. Once you're feeling comfortable about the mechanism and the campaign type, then is usually a time that you start to spend real money, and in the case that this is an organic campaign and you're not spending money, that money is probably people's time instead. I usually call that getting a campaign to production level, and if the campaign is going great, that's where you're gonna want to start to find the ceiling. If you've already proven everything out, you understand the dynamics of the campaign, the campaign has become predictable and repeatable, don't be afraid to go big. Go crazy big. Go ludicrously big. There should be no fear in this once you feel that you really gutturally understand how a campaign is working. An example from Datadog that we have is the biggest show in our industry is AWS, Amazon Web Services, re:Invent. That show is actually coming up In, I think, about 2 months, right after Thanksgiving, they're going to be doing it in person. We're going, once again, as big as we possibly can go. And since the first time that we attended, we've been the top—
sorry, can you go back?
We've been the top sponsor.
Can you go back a slide, please?
We've been the top sponsor. And for a long time, there was usually 4 top sponsors at re:Invent. One was Intel, big Fortune 500 company. One was Accenture, another one was Deloitte, really, really big consulting companies that have been around forever, no surprise there. And then there was a purple dog that nobody knew for years when we were a small startup. And we not only were getting the biggest spot, which was super intensive on the budget, but we had to man this ginormous booth space that we'd been given, and we would fly out literally for the first years, the entire company, and then portions of the company, and that translated to literally hundreds of people that we were flying out, putting up in Las Vegas for a week, and having them do all sorts of different tasks. But the thing was that re:Invent is the best campaign that we run. In every metric that we assess, it's actually our most efficient, we have access to the biggest customers that we end up bringing on board, and we power an entire year. Sometimes there are deals that are still about to be closed by the time that the next re:Invent rolls around that we've been working on because they're so large, for the entire year since we met them at the trade show. So for a long time, people would say like, "Oh, Datadog, what's up with you guys? You guys are bringing hundreds of people. There's purple shirts everywhere. You guys are crazy." Well, that's where we found our biggest customers and it's the best show, and if there were multiple re:Invents at the same investment of people, money, and time, I would love nothing better but to be able to invest over and over and over again, because once we figured out how that show worked, that's become our best campaign. A couple of words here on hiring. Once a campaign starts to feel like it's old hat, it's time to pass it on. Once you hit that ceiling and you, or maybe even before you hit the ceiling, you understand how this campaign type works or this channel works, time to pass it off, write up the playbook. Writing a playbook I think is a really good exercise also, 'cause when you start putting things on paper for somebody completely random to have to pick up and run with, you start to realize all of the gaps, all the holes, all the things that weren't actually fully thought out that you should put a little bit more time investment into to figure out before you're gonna get somebody completely net new to try to do this thing as well as you've been doing it for months or maybe even years. Hiring a specialist is of course always preferable, someone that's gonna bring back a background and experience in that particular area to grow things faster, or do things in a different direction, different from what you had thought to do, but sometimes an eager beaver, someone a bit more junior in their career that has a lot of enthusiasm and a lot of energy is a good match also, and just the sheer sweat equity can bring in a lot of new opportunities and new things that you wouldn't have thought to do and keeps on growing that program. Hiring an experienced leader is definitely a good shortcut for that learning curve. Someone can bring in a bag of tricks or a lot of experience of how this specific channel works that they can work on before or that have worked out for them before and you are able to take control of that channel faster. And I think a good rule of thumb is if you get to a point where you have 2 full-time employees working on that one campaign or that one campaign type, and especially if they're raising their hand and saying, I'm turning down good opportunities because I just don't have enough hours in the day to work on them. That's usually the time to bring in a leader to grow that market out or that marketing team out further and to try to take that team to grow explosively, to keep on finding the ceiling, and once they get there, maybe even broadening the scope of what they're working on to get more market opportunities. 'Cause at the end of the day, as companies, we're all out here for world domination. We want to become the next Salesforce.com, and the way to do that is to fully excavate a marketing channel up until you find the ceiling. So my last slide here, what happens when you've hit the promised land? You've scaled out a marketing channel, that pilot campaign from a couple months ago or maybe a couple of years ago is like just a pebble of what you're usually doing on an average day in terms of getting this big boulder pushed up the hill. I think one thing in particular is you can always do better, and I think that many people have heard the quote from Michael Jordan where he kept track of how many baskets he'd missed.
Sorry, can you go backwards?
You can always optimize. You can always find new opportunities. You can always keep on doing better. One important thing to note is I think that early on I thought, hey, once I get Google AdWords working great and I've got a great baseline, I can just take the day off. It'll be fine, no problem. But the thing is that everything around your campaigns is in constant flux. There's changes in platforms. There's changes and shifts in the market. There's changes in personas and what a person wanted last year is different from what they're going to want this year. Competitors are going to come in. Other products that support your ecosystem are going to come in. An edge is always going to erode. So once you've got your campaign set up great, you have to keep on taking a look at it, seeing what's going on, evaluating the numbers at a very fine scale. To make sure that your edge is staying as sharp as it was when you declared this thing.
Sorry, can you go back?
I don't know who's got the trigger finger. It's not me. So the edge is going to erode and you want to keep sharp. At certain inflection points, I think that this is for somebody that is getting to new levels of scale that they haven't worked on for that particular channel before, you're going to have to take apart your whole campaign because things are not gonna work as effectively or potentially at all anymore. Running a Google AdWords campaign that spends low thousands a month is different from a Google AdWords campaign that's spending millions a month, for instance. Having a 5x5 or a 10x10 booth is completely different from running a 30x30 monster, maybe even a double-decker booth with an upstairs where you're running meeting rooms. And when you take it apart, you're gonna potentially need new systems, new kinds of employees, new ways, a new playbook potentially, and then you're gonna have to put things back together to be functioning as effectively at that larger scale as you were at the smaller scale. And I think a very important note here, especially, I'm sure a lot of people here are in companies that are growing very quickly, and just like me, you're looking out into the market, hiring is getting tough, but once you're able to find your right hire, it's exciting, you wanna bring them in as quickly as possible. You're being told by your boards and by your management, grow, grow, grow, I think that team member number 10, and this is why this is the last thing I'm gonna talk about, I put it in red, is I think the flashing light for scaling a team where you need to maybe take a step back, slow down for a little bit, and similarly look at the team, analyze it, potentially take it apart and put it back together again. I have found that that 10th hire on a team is 2 or 3 hires away from getting to an inflection point where there's too many people coordinating, communicating becomes really difficult, And even though you've added way more people, the team is actually producing less because they haven't been built for scale. So I think that team member number 10 is the moment to stop. Document everything. Identify points that are ambiguous where you have to either document them very thoroughly, or you need to roll out a small training program for that team. Instituting regular check-in meetings, potentially coming in with more tooling so everybody can know what everybody else is working on. That's really the moment where if you're able to do that for people, you're gonna be able to keep on scaling because the thing is that for campaigns, they're relatively easy, I mean not easy, but they're straightforward is probably a better word, to peel back the onion, see what's going on, retool, maybe there's a very tailor-made solution that you plug in that system and the thing starts working well at scale again, but people are a little bit more challenging and to get them all working together can be more challenging, so that's why I definitely wanna highlight that that's probably the moment where you wanna stop and retool and make sure that you're built for scale. And building a team for more than 10 people, you're actually really building a team for 30 to 40 at that moment, and that's where you wanna be thinking of, because you also wanna be able to add people if you have found success at a large clip to where you quickly get to 30 or 40. So with that, Alright, now I do want to move to the next slide. Thank you everybody. Hopefully you've been able to glean some kind of insight or something to help you scale, and again, I think that the most important thing is to scale. With so many distractions, with so many things that marketing is being asked to do, if you don't get to that singular focus and you don't focus on that one thing, get the stats good, evaluate the stats, keep on doing repeated pilots, repeated enlargements of the campaign until you get to that scale and you try to find the ceiling, it's very difficult to get the subject matter expertise in that one channel to be able to hit that ceiling and get to the point where you can't do anymore. So with that, hopefully I'm able to help some of you guys hit that ceiling, and thank you very much.