Interview with Chris Hexton, co-founder and CEO of Vero – marketing software for the modern data stack
Discover Vero, the cutting-edge B2B marketing and customer communication solution that activates your Data Warehouse.
I had the privilege of interviewing Chris, the founder of Vero, a company that has gained recognition in the data industry for pioneering in the data-native email marketing space.
Vero was established in 2013 to create thermal marketing automation for software teams aspiring to enhance customer relationship processes. In 2022, the company introduced the world's first email marketing platform, Vero Connect, designed for the modern data stack. This platform was developed to assist data-driven businesses in connecting with their current user data and sending more cost-effective emails to more precisely segmented audiences. Vero Connect is designed to separate the "base" from the application, avoiding the complexity of traditional marketing software. Instead of requiring complex integrations and duplicate data storage, Vero Connect reads directly from your data warehouse, allowing you to query your user data and send personalized newsletters. This modular approach allows users to swap out their data architecture as their business grows. It builds with a vision of a world where the organization owns, stores, and manages customer data directly.
During our conversation, we discussed the evolution of email marketing tools, the potential of data-native applications that leverage data warehousing, and some general advice for those seeking to establish themselves in the modern data stack.
Can you walk us through your background and what led you to start Vero?
Yeah, for sure. We've been building Vero for over nine years now, so a long time. Before that, I worked in Pricewaterhouse Cooper's (PWC) IT risk department, where I gained some exposure to IT, but I was always a hobby programmer. I always played around with it in my teenage years, not building anything super complex.
My co-founder James and I have known each other since childhood and used to code together for fun during our university years. So James and I started a consulting company, and then we got exposure to customers through that consulting company, which exposed us to their problems.
And this is like 10 or 11 years ago now. Then, a problem we solved with the first version of Vero was that there wasn't a great tool to automate messages. There were great products at the high end, like Salesforce Marketing Cloud and Exact Target. And then there were tools like MailChimp, but they had no real automation. It's like fundamental stuff: if someone signs up for your product, say hello, educate them, share how to use the product best, that sort of thing. And so we thought there must be a gap between something simple and complex.
Does a consulting practice lead to this insight about its businesses' lack of automation?
Yeah. We were helping build software applications, generally web apps. And we built a few web apps, and as part of the brief, some of our customers asked us to include automation for activation and retention messages, but we couldn't find a tool to help us implement it. And so that's how we discovered the problem. Only a few customers had this problem, but it was enough to try and build a straightforward MVP and find some customers.
I need to follow the email marketing space closely. Can you explain how the industry has evolved during that ten years?
Yeah, absolutely. They are popular companies now, such as ActiveCampaign, that were around back then. Other tools had merged or been acquired: for example, Bronto was quite popular when we started for e-commerce companies, but I don’t see it very much these days. Sailthru and Campaign Monitor are still around, and I think they are part of the same broader company now as they've merged several businesses.
The main thing that changed around ten years ago was the birth of Segment.com. Segment has been a huge driving force behind the idea of “event tracking”. And, over the last decade, as the space has matured, the “next wave” of companies like Braze, CustomerIO, and ourselves have all been able to grow a lot.
It's been an exciting decade because you've got some companies that are steady and have been around for 20+ years and are still killing it, and there has also been a new wave of companies that have come up in the last decade, including Vero.
And I think the idea of behavioral real-time tracking trends has driven much of this new wave. That's certainly the wave that we've ridden.
Creating a universal email marketing platform is challenging. For example, the communication needs of B2B SaaS applications differ from those of large e-commerce companies. How do you approach that?
You're right; there's a different expectation regarding the data source. So most e-commerce companies are built on a platform like Shopify, or Magento, with standard customer behavior and backend. An e-commerce marketer doesn't have access to much of an engineering team because you don't necessarily need it to get started.This is a general statement, of course. As a result, generally, e-commerce businesses are looking for something they can just “click and connect” For example, connecting straight into Shopify. If you look at a company like Klaviyo, they focus on people building on top of Shopify and other out-of-the-box platforms.
Then you've got B2B software companies, and generally, the source of truth historically has been a CRM, Salesforce, HubSpot, whatever. Most well-established companies like Marketo and Autopilot, now called Ortto, are looking at that CRM space.
And then I think the area we've focused on is companies that are building their software. So that can be like mobile apps, games, web-based software, marketplaces, or any industry. This can also include e-commerce if they build or customize their backend. We have a great customer, a freight company, but they're a digital-first freight company. In all these companies, the data source is their product backend rather than a CRM or an off-the-shelf platform like Shopify. Like, that's the source of truth: what's in their product database.
Many MarTech tools charge per subscriber stored, which can be too expensive for most B2C companies and B2B product-led companies. However, if you are targeting sales-driven B2B companies that use a CRM as their primary data store, you need to charge more per subscriber because most B2B businesses have a smaller customer base. So the differentiator is how you make the price a better fit for a B2C company. You might make different technology choices or offer different options in terms of the basic feature set. At Vero, we focus on software companies, generally mid-market, and we focus on B2C and B2B product-led companies. Our goal is to provide a great value product that does the fundamentals really well for companies operating at scale.
Speaking about positioning, what is the persona in a company? Who are the main customers of Vero?
Currently, we have two versions of our product – Vero Connect and Vero Cloud. The product we've been selling for ten-plus years, Vero Cloud, has three personas in every company we sell to. There's the CMO or someone in charge of marketing or growth. There’s the person doing the marketing day-to-day, generally a digital marketer – they’re the person that will be logging into Vero the most, creating the content, all that sort of stuff. And the third persona is what you would now call a data engineer – an engineer responsible for getting data into the Vero platform.
Sometimes those three personas are the same person wearing all three hats. They've got the technical knowledge, can write code to content, and can buy the software. But as companies get bigger, each user tends to fall into one of the three buckets. And typically, for us, the head of marketing has the credit card and is the “buyer” persona.
When it comes to Vero Connect, the newer version of our product, we’ve observed that it is often data engineers driving the demand. It's usually a case of, you know, the company's been using a tool, let's say HubSpot, and they're getting frustrated with it for some reason. Maybe they're frustrated because the data in the tool is outdated or incomplete, and the marketers are trying to run some campaign, and they're like, “We can't trust this data,” or “We don’t have what we need.” Or maybe they're frustrated because the price is too high, as you said.
So the team is searching for a new solution, and a data engineer will say, “Hey, I saw Vero Connect somewhere—in a DBT Slack or whatever—we should add it to the list,” and then we will talk to the marketing team and show them that the interface is very familiar. Nothing different, nothing scary. And the data engineer champions the idea of a data warehouse first, and that's how we get in.
But that's a little different from the Vero Cloud product. Vero Cloud is generally more of a classic sale through marketing.
What influenced you to explore this idea when you decided to work in this Vero Connect?
We launched it in the second half of 2021, just over a year and a half ago. We launched Vero Connect because we’d heard direct feedback from customers and non-customers alike and wanted to fon something that we think is different and will be a growing market over the next five years. We were talking to people: current customers, customers of our competitors, etc., and, across about 90 conversations, we heard that the data warehouse was the most-trusted source of customer data. It's the source data teams like the most because there are a lot of tools to organize data in the warehouse, such as the wonderful “dbt’.
Yet it was annoying to get the data from your warehouse into Vero Cloud. And some customers said: why don't you just connect directly to the data warehouse? So we were like, well, let's try that. And that's why we launched the MVP that started Vero Connect.
So we did it deliberately, trying to find something where we could keep most of the product we had but narrow in on a niche and say, we're going to be the best for that thing over the next 3, 4, 5 years. We believe it's a trend that will become more popular, so we have the confidence to do it.
When when I first tried Vero Connect, I liked the idea that I could just write SQL queries to get my audience to craft segments. And speaking about digital marketers from your perspective, how tech-savvy are they, and do you feel any dynamic in terms of people becoming more familiar with SQL? Are you planning to make a low-code environment for them?
Yeah, it's an excellent question, and it's something we’re still solving. We've seen that most marketers, even of our core product, know what SQL is. And some of them are also happy to write queries with SQL, copy-paste, and make a few changes to “where” clauses and what-not.
But I would say that only a few are really comfortable with writing SQL from scratch. Particularly once you start doing “joins” or do some sort of windowing or aggregation: it can get tricky if you're unfamiliar with SQL.
Where we are trending at the moment is we intend to add features like “Saved audiences,” where you can save a SQL query and give it a name. The other thing we've had requests for is saving a SQL query with a name and adding variables so a marketer or a less technical user can alter these variable fields.
The third level, beyond the above two solutions, would be to introduce a full no-code SQL query builder—Metabase and Hightouch, and Census are doing similar things right now. That's a big undertaking, so we want to be sure that's the best way to do it.
Many of the Vero Connect customers at the moment want to create an environment where the product and marketing team have a lot of freedom to do their jobs but where there are some guardrails to ensure consistency.
As an example, guardrails around the design of emails. For example, I want to ensure my emails are on-brand. As a result, the most requests we've had so far are for this idea of “saved audiences” rather than a no-code audience builder. The data engineering team can create common SQL queries. And the marketers can just mix and match those. And, then, everyone has a hundred percent confidence that the definition of that group of customers is right.
One interesting thing about selling a marketing product is that most MarTech products try and sell marketers on the idea that they can do “whatever they want” without engineering help. If you go to most email marketing tools’ websites, they often promise the marketer the world.
And the reality is like that's never true in a modern organization, particularly a B2C or PLG B2B company where the product is core to the growth story. Suppose you want to do anything that's tightly integrated with the product experience. In that case, it will always be a collaboration between engineering and marketing, with both doing what they do best. Plus, as a company grows, it becomes a trickier thing because the marketers want to be free, but the other parts of the organization worry about QA and quality control.
One thing we’re excited about with Vero Connect is that this is a tool that data engineers like working with more than traditional MarTech solutions.
The modern stack has evolved over the last few years, and we see how we've been re-engineering existing things. For example, CDP has been introduced by Segment, but emerging tools like Hightouch or Census also call themselves CDP. Modern CRM systems like HubSpot and Snowflakes have a connection to the data warehouse and can also be classified as CDP. Every attempt to kill the CDP has made it stronger, has more people talking about it, and has more vendors claiming that they are. I'm trying to figure out Vero's place in this ecosystem and how to define our strategy.
I think the guiding light for us at the moment is to, which is partly driven by being a small team and the history of our product development, but you know, if you look at our core product, Vero Cloud, It has two parts to it. One is the CDP, the database we ran on behalf of our customers, and we gave them an API to store data in there. And then, the other half was the marketing interface layer, the UI on top of that. And with Vero Connect, we split that old product in half, and we're just selling the top part, and we're saying we'll sell you the UI layer, but you have to bring your data warehouse.
And so we're trying to be a more composable product. We now ourselves as good partners for Snowflake, RuddersStack, Snowplow, and other companies trying to be specialists at what they do. Our company can be a valuable partner for businesses that require email marketing and marketing technology tools. Unlike other tools that aim to be the central 360-degree view of the customer, we focus on doing our job well and charging less.
We are not wedded to collecting data: we are happy to leave that to other companies like Snowflake where it makes sense. Our goal is to be excellent partners. However, we do face challenges from indirect competitors like Census and HighTouch. While some customers may not notice the difference, others are finding it significant that we do not sync data at all.
The final thing I'd say is that because we're not storing the data, we can charge less than our competitors like Hubspot. And that's a deliberate part of the strategy as well. And for some of the people we're talking to, that's meaningful. If you've got 50 million users, we cut your bill in half or more, so that's pretty attractive.
How can we combine the data warehouse with event streams from different SaaS applications? Many companies have a data warehouse and use tools like Amplitude and Mixpanel for analytics. Are you considering merging these event streams to balance a single source of truth with the speed and flexibility of multiple sources?
Totally. That's exactly our plan. And if you summarize what I said before, my guiding strategy is to be the best solution for B2C and B2B product-led companies. And that involves being relatively low cost compared to our competitors and how we do that. By storing less data, that can also apply to the event tracking side.
And so we ultimately plan to merge our two products because that's what people have been requesting. They've been saying we want to query Snowflake when we build an audience for a newsletter. And sometimes, we also want to query it to kick off automation, but other times we want to trigger events or just make API calls to trigger automation because we need it in real-time.
And I think the difference we will make when we merge them is that, in the past, we've always stored all of the event histories forever because we're trying to be the gospel, the primary record of truth.
But, in the future, we'll just say the events might live for 30 days or however long you need them for the automation by default (we’ll give customers a choice to store for longer). After 30 days, we'll forget about them because we assume you're storing them in your data warehouse or somewhere else; you just need the real-time for the triggering so that that would be a difference, and that aligns with the strategy of not storing data, being low, cost, being a good fit for, for B2C, et cetera.
Are you planning to implement the connection between what you're doing in Vero Connect and the data warehouse to foster the feedback loop?
Yeah. We currently have webhooks so that you can use Fivetrain or Stitch, but we don't have that direct connection. It's a massive request that we write a copy of that back to the warehouse for any of the opens and clicks and all that sort of stuff. And it's pretty easy to do, given we already have the connection. It's on the roadmap for sure to do that. It makes a lot of sense for customers.
It makes me think about the direction where Snowflake is heading. Snowflake's direction is promising, especially with the announcement of Unistore, which combines OLTP data and analytics capabilities.
Yeah, what Snowflakes are doing is cool. One of the things from a data management point of view that's valuable about Vero Connect or building directly on top of the Datastore is that if you use a tool like HubSpot, it's straightforward to sync data to HubSpot. But if you want to audit the data, that's very hard because your marketing team can change it and don't write those changes back directly. And a few tools are trying to tackle this, such as Sequin.
But what's cool about Unistore and Snowpark and all the stuff Snowflake is doing, it will be possible for us to continue to build directly on Snowflake and even run transactional workloads for automation and push that worry off to Snowflake. And that means that the customer can maintain that central user table. And you wouldn't have to worry about auditing. So it removes the two-way data sync.
So yeah, it's fascinating because those transactional workloads exist. For us to make the automation side better and better, those databases will solve a key part of the problem. And if Snowflake does it, it should be a short time, and Google and Amazon will do it in some form too.
I was wondering how difficult it is to maintain connections with Snowflake, Redshift, BigQuery, etc., from an engineering perspective. How do you determine which connections to prioritize?
Yeah, I can go through as much as you know; know that I didn't write any of the code, but we've tried to build the tool so far to be agnostic or very flexible. In terms of supporting any data store, one of the strategic decisions is how much to invest in any data store.
For example, we were just talking about Snowflake with the Unistore. The Unistore will be something no other data store has. Should we build unique features that need that or not is a question?
Right now, we've decided to try and be as generic as possible. And so there's a little bit of thinking required in terms of different data stores like querying S3 and then querying Snowflake and even querying something like Airtable.
They're all going to have slightly different ways to batch-load an audience. Like in the case of Airtable, you have an iterate row by row. In the case of SQL, you can generally run one command to copy it to a CSV. You can stream from S3.
You know, they've all got slightly different access. So more just trying to architect things sensibly so it's easy to add a new data type. Once you get that part right, running the query is easy. The data warehouse does all the hard bits there. And then we just have some smarts around cashing it, validating emails, and de-duplicating everything.
I would like to know more about how our customers discover Vero. Data engineers seemed to be the main point of contact in the past, but I am curious if we have explored other avenues, such as marketing or cold outreach through email or LinkedIn.
At the moment, Vero Connect is still mostly being discovered via the data engineer, but Vero Cloud, our core business, is generally discovered by product marketers who read our blog Or who find us in the marketplaces of partners like Hightouch and Segment.com. Having been around for nine years, we get many eyeballs just from having written content for so long. We're also doing some of our outreach via email and LinkedIn.
Regarding persona, we try to talk to the marketer and the data engineer, depending on the business size and situation. Regarding Vero Connect specifically, we still need to figure out the product-market-channel fit — we’re learning with every sale.
We’re spending some time on partnerships with Snowflake. That'll be beneficial in the future. We're still a small company on their radar, so we have a long way to go. I think they seem like the most forward-thinking data warehouses at the moment, and they're smaller than AWS and Google, so it's a bit easier to build a relationship with the right people there (so far). Then there’s also manual outreach via email, LinkedIn, and whatnot. So yeah, we're getting there: it’s hard, hard work, as always, when you launch a new product!
And the question for you is, who is the brightest person in the marketing automation area? What does the internal operational and marketing stack look like in Vero?
What we do ourselves is: Use Segment (or RudderStack for any event tracking we do. I generally find that most useful for top-of-funnel stuff. Only some things need to have change data capture, and event tracking is like change data capture, so we mostly use that at our top-of-funnel marketing site tracking. We push all of our Segment data to Redshift, but, you know, Snowflake has some cool stuff that Redshift doesn't.
Then we plug Vero Connect directly on top of that, and away we go. I'm also a big fan of DBT. That's the only way to end up with some nice clean user and event tables and merge them in your data warehouse from various sources.
My knowledge of the Australian startup scene is limited, but I know some notable Australian B2B sales companies. However, my sample mainly comprises companies that typically bootstrap for the first few years, building a profitable business before seeking venture funding. My observations are biased, as I have only observed Canva, Atlassian, Deputy, and Vero. I am curious about what it is like to build a SaaS product in Australia.
Yep. Unfortunately, we have not been as successful as Canva and Deputy, but they're fantastic role models! For all of us in Australia. For all of us everywhere in the world, for that matter. There are three massive VC firms in Australia. And they were all founded about a decade ago. So to go and raise overseas, it was natural that companies took a bit longer to get there: it wasn't in their backyard. And then, as a result, people were slightly further along before they went overseas to raise their children.
Plus, everyone held up Atlassian as the poster child for bootstrapping for a long time. And, for better or worse, that rubbed off on a generation of Australian startups that took that too hard. That probably had a significant influence as well.
Do you have any staff outside of Australia, GTM folks, etc.?
Yeah. Most of our customers are outside Australia. So most of our product and engineering team is here, but most of our customer-facing teams are in the US and Europe.